— how it works5 min read

how to brief an agency properly.

a vague brief gets you a template deck. a sharp one gets you a real plan — and a comparable price.

— tl;dr

state the business problem, the numbers, the constraints, what you have tried, who decides, and the budget band. leave out the solution and the tactics — that is what you are paying them for.

agencies are not withholding their best thinking. they are pattern-matching against whatever you gave them, and most briefs give them a wish list rather than a problem. the difference between a generic proposal and a sharp one is almost entirely in the input. that is good news: it is the one part of the process you fully control, and it costs an afternoon.

this matters more than it used to. an agency reviewing four briefs a week will spend its senior hours on the two that are legible, and staff the rest from a template. a sharp brief is not politeness; it is how you get the good people put on your account before you have paid anyone anything.

— 01the six parts.

this fits on one page. anything longer and the important parts get skimmed.

the order is deliberate: problem before numbers, numbers before constraints. agencies read briefs looking for the thing they are being hired to change, and if that appears on page two under a heading called background, the proposal will be built around whatever appeared on page one instead. put the uncomfortable sentence first — the one that names what is not working.

be specific about who this is for, too. "everyone in the uae" is not an audience, and it produces work aimed at nobody. one or two real customer descriptions — job, city, what they were doing five minutes before they found you — will change a proposal more than any amount of brand adjective.

include the awkward context too. a product that is more expensive than every competitor, a founder who will not appear on camera, a sales team that takes three days to call back — these shape the work far more than anything in a brand deck, and hiding them only guarantees a proposal you will have to renegotiate in month two.

— the one-page brief
1
the business problem
not "we need more content". "we get 40 enquiries a month and need 100 to hit the plan."
2
the numbers you have
traffic, enquiries, close rate, average order value, current spend. rough is fine; absent is not.
3
constraints
markets, languages, compliance rules, what the founder will and will not appear in.
4
what you have tried
and what it cost. this single section saves an agency a month of rediscovery.
5
who decides
one named person, and how fast they approve. this predicts your first month more than anything else.
6
the budget band
a range, not a secret. it makes three proposals comparable instead of three guesses.

— 02why you should name the budget.

the most common objection is that naming a budget invites the agency to spend all of it. in practice, withholding it wastes everyone's time: you get three proposals at three unrelated scales and cannot compare any of them. a band is enough — "we are thinking somewhere between x and y a month" lets a good agency tell you honestly what that buys and what it does not.

it also filters. an agency whose floor is well above your ceiling will say so in a day, which saves you three meetings.

if you genuinely do not know the number, put a decision rule in instead: "we will spend what returns a customer under six months of payback." that gives a competent agency enough to propose a starting scope and a scaling trigger, which is a more useful conversation than a fixed monthly fee anyway.

— 03what to leave out.

two things. first, the solution — "we need twelve reels a month and a tiktok account" is a tactic dressed as a brief, and it prevents the agency from telling you that your landing page is the actual problem. describe the outcome and let them argue for the method.

second, the fake urgency. a deadline that is not real trains the relationship to treat deadlines as negotiable, and it usually produces a rushed first month you then have to redo.

leave out the competitor screenshots as well, or at least demote them. "we want to look like them" is the fastest route to work that is one step behind a rival, and it tells the agency to imitate rather than position. reference is useful; instruction is not.

one last discipline: write the brief for someone who does not work in your industry. jargon hides thinking, and a brief full of internal shorthand will get a proposal full of agency shorthand in return. if a friend outside your sector can read it and tell you what business problem you are trying to solve, it is finished. if they cannot, the agencies will guess — and their guesses will all be different, which is how three incomparable proposals end up on your desk.

— 04what a good response looks like.

a proposal that quotes your own numbers back at you, disagrees with at least one thing in the brief, names the people who would work on it, and states what happens in month one as deliverables rather than adjectives. if all three responses look interchangeable, the brief was the problem — not the agencies.

give them a week, and take a question call in the middle of it. the questions an agency asks are more diagnostic than the deck it produces: someone asking about your close rate and your sales process is thinking about outcomes, and someone asking only about brand guidelines and posting frequency is thinking about output. if the brief is really about how you look rather than what you sell, say so — that is a digital branding agency job and it should be scoped as one.

read next: how to brief a content creator, how to brief a shoot, how to brief a videographer in dubai.

— the short version
brief the problem, the numbers and the budget band — not the tactics. then judge the response by whether it disagrees with you. send us a brief →
frequently asked.
should we send an rfp to several agencies?
two or three is useful. more than that and you get a race to look impressive rather than to be right, and you will not have time to interrogate any of them properly.
should we tell agencies our budget?
yes, as a band. it is the only way to compare proposals, and it filters out anyone whose floor is above your ceiling.
how long should a brief be?
one page. six sections. if it runs to ten pages, the important parts will be skimmed.
briefingprocessagencies
D
— written by
Deepika
Social Mafia

part of the studio team across dubai and mohali.